Bit Error Ratio (BER) is a metric used to evaluate transmission system quality in communications equipment. Communications equipment may be manufactured with a target BER, referring to the BER under the worst-case noise conditions. Since it is desirable for equipment to be manufactured with a noise margin, which is the additional noise required beyond a nominal noise power level to reach the target BER, it is desirable for equipment to be tested in noise conditions that have noise greater than that in the worst-case operating noise environment.
In conventional controlled tests of communications equipment, a data signal is transmitted from a piece of communications equipment and a noise signal may be added to the data signal prior to reception to generate a noisy signal. The noisy signal may be received and the BER measured. The process of adding the noise signal in such a scenario may change the channel seen by the data signal relative to what the equipment would experience in the field. For relatively low speed data, the changes to the channel caused by controlled tests have a negligible effect on the BER as compared with the additive noise.
However, it may not be practical to measure the effects of additive noise in very high speed systems, e.g., systems faster than several gigabits per second (Gbps), and thus, the test scenario described above for testing noise margin has not been popular in the industry for very high speed systems. This is mainly because any noise injection circuitry itself causes additional impairments to the channel such as increasing loss, reflections, and uncontrolled crosstalk and Electromagnetic Interference (EMI). The BER due to these impairments may be significant relative to the BER due to additive noise. Also, an accurate noise margin test is difficult because any added noise level is not easy to control. And finally, there may be no physical space available for auxiliary devices for controlled noise generation once equipment is built for high-density backplanes. Despite these difficulties, there have been noise tolerance tests conducted in laboratories on specially built test boards, but significant differences between the noise environments of the test boards in a lab and the actual deployed systems in the field have rendered the results from such lab testing ineffective for use in accurate product qualifications. Therefore, there is a need for efficient and accurate noise margin tests for systems with data rates over several Gbps.